|Publication number||US6927473 B2|
|Application number||US 10/620,054|
|Publication date||Aug 9, 2005|
|Filing date||Jul 14, 2003|
|Priority date||Oct 2, 1998|
|Also published as||US6277674, US6511868, US6703263, US7109105, US7816246, US20010046726, US20030134456, US20040036090, US20040209404, US20060270208|
|Publication number||10620054, 620054, US 6927473 B2, US 6927473B2, US-B2-6927473, US6927473 B2, US6927473B2|
|Inventors||Zhongze Wang, Michael P. Violette, Jigish Trivedi|
|Original Assignee||Micron Technology, Inc.|
|Export Citation||BiBTeX, EndNote, RefMan|
|Patent Citations (13), Non-Patent Citations (1), Referenced by (4), Classifications (11), Legal Events (8)|
|External Links: USPTO, USPTO Assignment, Espacenet|
This application is a divisional of application Ser. No. 10/331,033, filed Dec. 27, 2002, now U.S. Pat. No. 6,703,263, issued Mar. 9, 2004, which is a continuation of application Ser. No. 09/901,761, filed Jul. 9, 2001, now U.S. Pat. No. 6,511,868, issued Jan. 28, 2003, which is a continuation of application Ser. No. 09/165,754, filed Oct. 2, 1998, now U.S. Pat. No. 6,277,674, issued Aug. 21, 2001.
The present invention relates generally to integrated circuit design and fabrication. Specifically, the present invention relates to semiconductor fuses, methods for fabricating the semiconductor fuses, methods for using the semiconductor fuses, and semiconductor devices containing the semiconductor fuses.
Computers typically have various types of devices which store data, such as memory devices. One type of memory device is a read only memory (ROM) device in which data is permanently stored and cannot be overwritten or otherwise altered. Thus, ROM devices are useful whenever unalterable data or instructions are required. ROM devices are also non-volatile devices, meaning that the data is not destroyed when power is shut off. ROM devices are typically programmed during fabrication by making permanent electrical connections in selected portions of the memory device. One disadvantage of ROM devices is that their programming is permanently determined during fabrication and, therefore, can only be changed by redesign.
Another type of memory device is a programmable read only memory (PROM) device. Unlike ROM devices, PROM devices are programmable after their design and fabrication. To render them programmable, PROM devices are typically provided with an electrical connection in the form of a fusible link (fuse). There are a considerable number of fuse designs used in PROM devices, such as those disclosed in IEEE Transactions on Electron Devices, Vol. 33, No. 2, p. 250-253 (February 1986), and in U.S. Pat. Nos. 5,589,706, 4,491,860, 5,625,218, 4,796,075, and 4,740,485, the disclosures of each of which are incorporated herein by reference. Perhaps the most common fuse design is a metal or polysilicon layer which is narrowed or “necked down” in one region. To blow the fuse, a relatively high electrical current is driven though the metal or polysilicon layer. The current heats the metal or polysilicon above its melting point, thereby breaking the conductive link and making the metal layer or polysilicon discontinuous. Usually, the conductive link breaks in the narrowed region because the current density (and temperature) is highest in that region. The PROM device is thus programmed to conducting and non-conducting patterns, thereby forming the 1 or 0 comprising the data stored in the memory device.
Rather than employing an electrical current, a laser can be employed to blow the fuses. Using lasers instead of electrical current to blow fuses, however, has become more difficult as the size of memory devices decreases. As memory devices decrease in size and the degree of integration increases, the critical dimensions (e.g., fuse pitch) of memory cells become smaller. The availability of lasers suitable to blow the fuse becomes limited since the diameter of the laser beam should not be smaller than the fuse pitch. Thus, the fuse pitch, and the size of memory devices, becomes dictated by the minimum diameter of laser beams obtainable by current laser technology.
The ability of electrical currents to blow fuses could aid in adapting fuses for a variety of applications, such as redundancy technology. Redundancy technology improves the fabrication yield of high-density memory devices, such as SRAM and DRAM devices, by replacing failed memory cells with spare ones using redundant circuitry, which is activated by blowing fuses. Using laser beams to blow the fuses limits the size and, therefore, the number of memory devices as explained above since the diameter of a conventional laser beam is about 5 microns. Using electrical currents instead to blow fuses, therefore, has a greater potential for high-degree integration and decreased size of memory devices.
The present invention generally provides fuses for integrated circuits and semiconductor devices, methods for making the same, methods of using the same, and semiconductor devices containing the same. The semiconductor fuse of the present invention contains two conductive layers, an overlying layer and an underlying layer, disposed on an insulating substrate. The underlying layer comprises a refractory metal nitride, such as titanium nitride, and the overlying layer comprises tungsten silicide. The semiconductor fuse may be fabricated during manufacture of local interconnect structures containing the same materials.
The present invention includes a semiconductor fuse comprising an insulating substrate, a titanium nitride layer disposed over the insulating substrate, and a tungsten silicide layer disposed over the titanium nitride layer. The insulating substrate maybe an isolation region disposed on a silicon or other semiconductor substrate. The titanium nitride layer and the tungsten silicide layer may have a similar pattern, which includes a neck portion located between terminal portions. The neck portion may be smaller in width than the terminal portions and may have a width of about 0.35 microns and length of about 3.5 microns. The inventive semiconductor fuse may be contained in an integrated circuit, either alone or with a local interconnect structure.
The present invention also includes a method of making a semiconductor fuse by providing an insulating substrate, forming a titanium nitride layer over the insulating substrate, and forming a tungsten silicide layer over the titanium nitride layer. The insulating substrate may be a field oxide region formed by thermally oxidizing a portion of a silicon substrate. The titanium nitride layer may be formed by depositing a layer of titanium and annealing the titanium in an atmosphere containing nitrogen or by depositing a layer of titanium in an atmosphere containing nitrogen. The tungsten silicide layer may be formed by chemical vapor deposition. The tungsten silicide layer may be patterned by a photolithographic pattern and etch process and the titanium nitride layer may then be patterned by a wet etch process using the patterned tungsten silicide layer as a hard mask. The method for making the semiconductor fuse can be incorporated into a method for making an integrated circuit containing the fuse alone or containing the fuse and a local interconnect structure.
The present invention also includes a method of using a semiconductor fuse by first providing a semiconductor fuse comprising a tungsten silicide layer and a titanium nitride layer disposed on an insulating substrate and having a neck portion located between terminal portions and then flowing sufficient electrical current to blow the fuse by causing the neck portion of the tungsten silicide layer to melt. When the neck portion has a width of about 0.35 microns and a length of about 3.5 microns, the electrical current sufficient to blow the fuse is about 5.5 mA and the leakage current of the blown fuse is about 1 nA.
By fabricating the semiconductor fuse with tungsten silicide and titanium nitride over an insulating substrate, the fuse of the present invention can be manufactured while also fabricating a local interconnect structure with the same materials. The inventive semiconductor fuse, which is used to program redundant circuitry, can be blown by electrical current rather than laser beams, thus allowing the fuse pitch to be smaller than the pitch of prior art fuses which are blown by laser beams. Another advantage of the inventive semiconductor fuse is that the fuse may be blown by less electrical current, on the order of less than half of the current required to blow conventional polysilicon fuses.
The Figures presented in conjunction with this description are not actual views of any particular portion of an actual semiconductor device or component, but are merely representations employed to more clearly and fully depict the present invention.
The following description provides specific details, such as material thicknesses and types, in order to provide a thorough understanding of the present invention. The skilled artisan, however, would understand that the present invention may be practiced without employing these specific details. Indeed, the present invention can be practiced in conjunction with fabrication techniques conventionally used in the industry.
The process steps and structures described below do not form a complete process flow for manufacturing integrated circuit (IC) devices or a completed device. Only the process steps and structures necessary to understand the present invention are described below.
As shown in
Pad oxide layer 4, which acts as a stress relief layer, is then formed over the upper surface of substrate 2 by any suitable process known in the art. Pad oxide layer 4 may be thermally grown or deposited by any suitable process known in the art, such as chemical vapor deposition (CVD) to a thickness of about 50 to about 200 angstroms. Preferably, pad oxide layer 4 comprises silicon oxide formed by thermal oxidation of the preferred silicon substrate 2.
Silicon nitride layer 8 may then be deposited over pad oxide layer 4. Any suitable process known in the art, such as a CVD process, can be employed to deposit silicon nitride layer 8 to a thickness ranging from about 50 to about 2000 Å. As explained below, silicon nitride layer 8 serves as a mask when forming isolation regions 10 (FIG. 2).
Next, as shown in
Next, as depicted in
Next, as depicted in
Next, polysilicon layer 14 is formed over dielectric layer 12 and field isolation regions 10. Polysilicon layer 14 may be formed by any suitable deposition method known in the art, such as physical or chemical vapor deposition. Preferably, polysilicon layer 14 is deposited by low-pressure CVD to a thickness ranging from about 800 to about 2000 angstroms. Polysilicon layer 14 is then doped, preferably with an n-type dopant. such as phosphorous, by any suitable ion implantation or doping process known in the art. Alternatively, polysilicon layer 14 can be in-situ doped during deposition of polysilicon layer 14 by including a gas containing the desired dopant in the deposition atmosphere.
Next, conductive layer 16 is formed over polysilicon layer 14 by any suitable process known in the art. Conductive layer 16 may comprise any conductive material that has a lower resistance and/or a lower melting point than polysilicon layer 14, such as titanium suicide, polycide, or tungsten silicide. Preferably, tungsten suicide is employed as conductive layer 16. When tungsten silicide is employed as conductive layer 16, the tungsten silicide layer may be formed by any process yielding the desired physical and chemical characteristics, such as CVD or co sputtering. Preferably, this tungsten suicide layer is formed by CVD using tungsten hexafluoride (WF6) and silane (SH4) at a temperature ranging from about 400 to about 500° C. until a thickness of about 600 to about 1500 angstrorns is obtain.
Next, as depicted in
If desired, sidewall spacers 21 for gate structure 20 can then be formed by any suitable process known in the art, such as depositing a dielectric layer overall and etching anisotropically to leave substantially vertical sidewall spacers 21. At least one diffusion region (illustrated as diffusion region 6), such as source/drain regions, can then be formed in substrate 2 by any suitable process known in the art, such as implanting a suitable dopant at an energy and dose sufficient to form the desired dopant concentration. For example, arsenic could be implanted at about 20 keV to about 80 keV to obtain a concentration of about 1×1020 to about 5×1020 atom/cm3.
The Ti layer is then converted to a layer comprising titanium and nitrogen, such as TixNy where x can range from more than 0 to less than 1.0 (hereafter “titanium nitride (or TiN) layer 24”. In one embodiment, this conversion is performed by annealing the Ti layer in a nitrogen-containing atmosphere for a time and temperature sufficient to convert the titanium to a mixture of titanium and nitrogen. In this annealing process, the temperature may range from about 600 to about 750° C., and is preferably about 650° C., and the time may range from about 20 to about 120 seconds, and is preferably about 60 seconds. The nitrogen-containing atmosphere of the annealing process may comprise a gas or a mixture of gases containing nitrogen, such as nitrogen, ammonia, or mixtures thereof. The annealing atmosphere may also contain other gases, such as argon or hydrogen. Preferably, the nitrogen-containing atmosphere contains substantially pure nitrogen gas.
TiN layer 24 may also be formed by an alternative process. In this process, TiN layer 24 is formed in a single step by depositing Ti over substrate 2, isolation regions 10, and gate structure 20 in an atmosphere comprising N. Preferably, during this single step process, the Ti is deposited by any suitable deposition process, such as sputter deposition using a Ti target in a vacuum containing Ar and N. TiN layer 24 need not be pure titanium nitride. The concentration of nitrogen in TiN layer 24 can be controlled by controlling the nitrogen concentration in the gas or gases comprising the sputtering atmosphere. For example, the concentration of nitrogen in TiN layer 24 may range from about 2% to about 80%, and preferably about 10%.
As shown in
Next, conductive layer 26 is patterned in the desired fuse pattern. This patterning can be performed by any suitable process known in the art, such as a photolithographic pattern and etch process. For example, a photoresist film can be spun on conductive layer 26, developed, and portions thereof removed to leave photoresist mask 23 (shown by the dotted line in FIG. 5). Using photoresist mask 23, portions of conductive layer 26 not underlying photoresist mask 23 are removed by any suitable anisotropic etching process known in the art.
Removing portions of conductive layer 26 will expose underlying portions of TiN layer 24, which are then also removed. Any suitable process which removes portions of TiN layer 24 without removing or adversely affecting the rest of the underlying device can be employed. Preferably, a wet etch solution is employed. More preferably, a wet etch solution containing about 5 parts H2O, about 0.25 to about 1 part H2O2, and about 0.25 to about 1 part NH4OH is used to remove the desired portions of TiN layer 24. After removal of the underlying portions of TiN layer 24, the structure of the IC containing the fuse and local interconnect (LI) is depicted in FIG. 7.
This process of patterning conductive layer 26 and TiN layer 24 should be performed to obtain the desired LI structure 36 and to obtain the desired structure of fuse 34. One preferred structure of an LI structure 36 is illustrated in
Neck portion 30 is that portion of fuse 34 that will blow when subjected to programming or sufficient electrical current. For the preferred dimensions of neck portion 30 above, when a sufficient amount of current, about 1 to about 25 mA and preferably about 5.5 mA, flows through conductive layer 26, it heats up and melts in neck portion 30, thereby interrupting the current flow. Neck portion 30 blows before terminal portions 28 because, while the same amount of current runs through both, there is less area in neck portion 30. Consequently, the temperature of neck portion 30 is higher than the temperature in terminal portions 28, leading to quicker melting of conductive layer 26 in this region. Reducing the width-to-length ratio of neck portion 30 and changing the material of conductive layer 26 will change the amount of current needed to blow the fuse 34. Tungsten suicide is the preferred material for conductive layer 26 since, when practiced of the present invention, the tungsten suicide requires only about half the electrical current to blow as a polysilicon fuse with similar dimensions. After the fuse is blown by this electrical current, the leakage current of the blown fuse ranges from about 1 to about 10 nA and is preferably less than about 1 nA.
Further processing of the IC can then proceed. An optional high temperature anneal may be performed to reduce the local interconnect sheet resistivity. This high temperature anneal may be performed at 750-875° C. for about 15 to about 60 minutes in an argon atmosphere. Metal contacts could then be formed to terminal regions 28 of fuse 34 by any suitable process known in the art.
Further enhancements to the above disclosed method are possible. For example, a dielectric layer could be formed over conductive layer 26 in gate structure 20, if desired. Also, the fuse of the present invention could be fabricated simultaneously with devices other than local interconnects.
Having thus described in detail the preferred embodiments of the present invention, it is to be understood that the invention defined by the appended claims is not to be limited by particular details set forth in the above description, as many apparent variations thereof are possible without departing from the spirit or scope thereof.
|Cited Patent||Filing date||Publication date||Applicant||Title|
|US4491860||Apr 23, 1982||Jan 1, 1985||Signetics Corporation||TiW2 N Fusible links in semiconductor integrated circuits|
|US4740485||Jul 22, 1986||Apr 26, 1988||Monolithic Memories, Inc.||Method for forming a fuse|
|US4783424||Feb 21, 1986||Nov 8, 1988||Tokyo Shibaura Denki Kabushiki Kaisha||Method of making a semiconductor device involving simultaneous connection and disconnection|
|US4796075||Apr 15, 1986||Jan 3, 1989||Advanced Micro Devices, Inc.||Fusible link structure for integrated circuits|
|US5589706||May 31, 1995||Dec 31, 1996||International Business Machines Corp.||Fuse link structures through the addition of dummy structures|
|US5618750||Apr 13, 1995||Apr 8, 1997||Texas Instruments Incorporated||Method of making fuse with non-corrosive termination of corrosive fuse material|
|US5625218||Jun 16, 1995||Apr 29, 1997||Nippondenso Co., Ltd.||Semiconductor device equipped with a heat-fusible thin film resistor and production method thereof|
|US5914524||May 12, 1995||Jun 22, 1999||Kabushiki Kaisha Toshiba||Semiconductor device|
|US5990537||Dec 15, 1993||Nov 23, 1999||Fujitsu Limited||Semiconductor device with fuse|
|US6163062 *||May 4, 1998||Dec 19, 2000||Mitsubishi Denki Kabushiki Kaisha||Semiconductor device having a metallic fuse member and cutting method thereof with laser light|
|US6261937 *||Jun 24, 1998||Jul 17, 2001||Siemens Aktiengesellschaft||Method for forming a semiconductor fuse|
|US6277674||Oct 2, 1998||Aug 21, 2001||Micron Technology, Inc.||Semiconductor fuses, methods of using the same, methods of making the same, and semiconductor devices containing the same|
|US6511868 *||Jul 9, 2001||Jan 28, 2003||Micron Technology, Inc.||Semiconductor fuses, methods of using the same, methods of making the same, and semiconductor devices containing the same|
|1||Fukuda, Yukio, et al., A New Fusible-Type Programmable Element Composed of Aluminum and Polysilicon. IEEE Transactions on Electron Devices, vol. ED-33, No. 2, Feb. 1986.|
|Citing Patent||Filing date||Publication date||Applicant||Title|
|US7829991||Oct 18, 2007||Nov 9, 2010||Micron Technology, Inc.||Stackable ceramic FBGA for high thermal applications|
|US8072082||May 28, 2008||Dec 6, 2011||Micron Technology, Inc.||Pre-encapsulated cavity interposer|
|US8367494||Apr 5, 2011||Feb 5, 2013||International Business Machines Corporation||Electrical fuse formed by replacement metal gate process|
|US8399297||Oct 20, 2011||Mar 19, 2013||Micron Technology, Inc.||Methods of forming and assembling pre-encapsulated assemblies and of forming associated semiconductor device packages|
|U.S. Classification||257/529, 257/352, 257/E23.15|
|Cooperative Classification||H01L21/76888, H01L23/5256, H01L23/5258, H01L2924/0002|
|European Classification||H01L21/768C8B, H01L23/525F, H01L23/525F4|
|Jan 7, 2009||FPAY||Fee payment|
Year of fee payment: 4
|Jan 9, 2013||FPAY||Fee payment|
Year of fee payment: 8
|May 12, 2016||AS||Assignment|
Owner name: U.S. BANK NATIONAL ASSOCIATION, AS COLLATERAL AGEN
Free format text: SECURITY INTEREST;ASSIGNOR:MICRON TECHNOLOGY, INC.;REEL/FRAME:038669/0001
Effective date: 20160426
|Jun 2, 2016||AS||Assignment|
Owner name: MORGAN STANLEY SENIOR FUNDING, INC., AS COLLATERAL
Free format text: PATENT SECURITY AGREEMENT;ASSIGNOR:MICRON TECHNOLOGY, INC.;REEL/FRAME:038954/0001
Effective date: 20160426
|Mar 17, 2017||REMI||Maintenance fee reminder mailed|
|Jun 8, 2017||AS||Assignment|
Owner name: U.S. BANK NATIONAL ASSOCIATION, AS COLLATERAL AGEN
Free format text: CORRECTIVE ASSIGNMENT TO CORRECT THE REPLACE ERRONEOUSLY FILED PATENT #7358718 WITH THE CORRECT PATENT #7358178 PREVIOUSLY RECORDED ON REEL 038669 FRAME 0001. ASSIGNOR(S) HEREBY CONFIRMS THE SECURITY INTEREST;ASSIGNOR:MICRON TECHNOLOGY, INC.;REEL/FRAME:043079/0001
Effective date: 20160426
|Sep 4, 2017||LAPS||Lapse for failure to pay maintenance fees|
Free format text: PATENT EXPIRED FOR FAILURE TO PAY MAINTENANCE FEES (ORIGINAL EVENT CODE: EXP.)
|Sep 26, 2017||FP||Expired due to failure to pay maintenance fee|
Effective date: 20170809